Research and analysis

Pension Credit ‘Invitation to Claim’ Trial

Updated 24 October 2024

Quantitative and qualitative research with claimants

Clara Yates, Gary Gifford, Catriona Parpworth and Joe Shuter Department for Work & Pensions

DWP ad hoc research report no. RR 96

A report of research carried out by the Department for Work and Pensions

Crown copyright 2024

If you would like to know more about DWP research, email [email protected]

First published October 2024

ISBN 978-1-78659-718-2

Executive summary

Background

Pension Credit (PC) provides pensioner households with extra money to help with living costs if they are over State Pension age (SPa) and on a low income. The latest take-up estimates indicate that in the financial year ending 2022, 63% of those entitled to PC received the benefit^{[footnote 1]}. This means that there are a significant number of pensioner households that are currently not claiming the benefit they could be entitled to.

While these estimates provide a useful starting point for thinking about take-up, it remains difficult to accurately identify PC entitled non-recipients (ENRs). Results from the Family Resources Survey (FRS) can, for example, reveal savings and income data at an aggregate level, but cannot provide data at the individual household level. Without this data, we are unable to contact all potential PC ENRs and cannot target take-up activity effectively.

To ensure that pensioners receive the support they are entitled to, the Department for Work and Pensions (DWP) ran a test and learn exercise, exploring a new way of using data more effectively. The project involved using Housing Benefit (HB) data to identify a subset of all known PC ENRs^{[footnote 2]} and sending them letters, inviting them to make a claim.

Research aims

This research was conducted with the aims of:

1. Identifying whether DWP could use administrative HB data to accurately identify PC ENRs;

2. To investigate the effectiveness of sending an invitation to claim letter to a sample of those identified from HB data as being entitled to PC, but not receiving the benefit.

Research design

HB administrative data was used to identify approximately 144,500 pensioner households who were potentially entitled to PC but not receiving it. A treatment group of 2,409 pensioner households within 10 Local Authority (LA) areas were sent a letter in July 2023, advising them of their potential eligibility and inviting them to make a claim. They also received a reminder letter in September 2023. The remaining approximately 142,000 pensioner households outside of the 10 LA areas, did not receive a letter and were treated as a control group for comparative purposes.

Administrative data was used to track PC claims made subsequently to the invitation to claim letter. Early claims data provided essential learning and allowed us to improve the criteria used to identify ENRs with HB data. Follow up interviews were conducted with a number of pensioner households from the treatment group.

Interviews covered the participant’s claims history; reactions and understanding of the letter, and reasons to claim or not to claim.

Key findings

• A significant outcome of this test and learn exercise is that we have been able to improve the ENR selection criteria. Data identified during the test and learn exercise allowed us to refine the criteria to more accurately identify potential ENRs. This criteria was applied to the original treatment group, producing a smaller sub-group (the refined group) that we can be more confident are ENRs. The results of the refined group are more indicative of the success of this exercise, however for completeness the results of the original group are also reported.

• Applying the original criteria:

  • 713 (29.6%) of the 2,409 households that were sent a letter made a PC claim during the almost 4-month period following the mailing of the letters.

  • DWP assessed the eligibility of these 713 claims and found that 267 (37.4%) were entitled to a mean average of just over £46 per week of PC.

  • This means that 11.1% of those that DWP wrote to made a successful PC claim, the comparative figure for the control group was 2.3%.

• Applying the refined criteria reduced the treatment group to 708.

  • Of these, 259 (36.6%) made a claim to PC in the almost 4-month period following the mailing of the letter.

  • When assessed, 195 (75.3%) of these 259 claims were found to be entitled to a mean average of just under £51 per week of PC.

  • This means that when applying the refined criteria, 27.5% of those DWP wrote to in the refined criteria subsequently made a successful PC claim.

  • When applying the refined criteria to the control group also, the comparative figure was only 4.6%.

  • These findings suggest that sending an ‘Invitation to Claim’ PC letter to ENRs significantly increased the likelihood of successfully making a claim, with those in the refined treatment group being 6 times more likely than those in the refined control group to make a successful PC claim. Receipt of the letter also had a noticeable impact on those who were entitled to smaller amounts of PC, significantly increasing their likelihood of making a successful claim.

These findings are illustrated in the figure below^{[footnote 3]}

Original treatment group

Step 1: Sent an invitation to claim letter (n=2,409 households)

Step 2: 713 households made a claim after the invitation to claim letter

Step 3: 267 households from the original treatment group made a successful claim

Refined treatment group

Step 1: The refined treatment group (n=708 households) is a subset of the original treatment group

Step 2: 259 household from the refined treatment group made a claim after the invitation to claim letter

Step 3: 195 households from the refined treatment group made a successful claim

• 19 qualitative telephone interviews were conducted with individuals who received the letter.

  • 15 participants said they made a claim after receiving the invitation to claim letter. Of those who were asked, all said that the letter had influenced their decision to claim. Secondary motivations to make a claim included believing it was ‘worth a try’ (worth making a claim to see if they were entitled) and struggling with the cost of living.

  • Generally understanding of the letter was good, with most participants interpreting the invitation to claim letter as meaning they were potentially entitled to PC and encouraging them to make a claim.

• Overall, feedback on the invitation to claim letter was positive. Even those who claimed and were found to be ineligible appreciated DWP taking a lead in encouraging ENRs to claim money they are potentially entitled to.

• Assuming the results from the refined group were replicated to apply to the whole of the estimated population sample, it could result in a further:

  • 32,000 (28%) to 41,000 (35%) making a successful claim to PC;

  • 1 to 2 percentage point increase in PC take-up.

Acknowledgements

This research was conducted in-house by analysts in the Pensions and Later Life Analysis Division at DWP.

We are grateful for the guidance and support offered by DWP colleagues as part of this work.

We would like to thank all of our research participants for giving up their time to take part in the telephone interviews, and for sharing their experiences and insight.

Authors

This report was written by Clara Yates, Gary Gifford, Catriona Parpworth and Joe Shuter.

The authors are analysts in the Pensions and Later Life Analysis Division at DWP.

1. Introduction

 1.1  Background

1.1.1 Pension Credit

PC provides those over SPa and on a low-income extra money to help with living costs.

PC guarantees that no-one above SPa need live on an income of less than a guaranteed amount. These amounts are more if pensioners have caring responsibilities, are severely disabled or have housing costs. The Guarantee Credit (GC) works by topping up pensioners’ weekly income to this level. The second element is the Savings Credit (SC) which is available for pensioners who reached SPa before 6 April 2016. Some people will be entitled to GC, SC or both.

1.1.2 Pension Credit take-up

Estimates of PC take-up indicate that many people who DWP would expect to be entitled to the benefit do not make a claim. The latest estimates of take-up were published in October 2023^{[footnote 4]}. These figures are based on DWP administrative data, Local Authority (LA) administrative data, and data from the FRS.

The estimates showed that in the financial year ending 2022:

• 6 out of 10 of those entitled to PC claimed the benefit.

• Up to 880,000 families who were entitled to receive PC did not claim the benefit.

• 73% of the total amount of PC available was claimed.

• Up to £2.1 billion of available PC went unclaimed.

• On average, this amounted to around £2,200 per year for each family entitled to receive PC who did not claim the benefit.

Research was conducted in 2012 to explore the barriers to claiming PC among ENRs^{[footnote 5]}. This study used as its sample 2,224 people who were identified as being entitled to PC. The results showed that:

• While nearly three quarters of ENRs had heard of PC, among this group 54% had only heard the name and only 16% said they were well informed about PC.

• Participants had limited knowledge of who to contact if they required information about the State Pension or other benefits.

• The primary reason given for not claiming was perceived ineligibility with 65% of ENR participants saying that they did not claim because they did not think they were entitled / no longer entitled / have too much money.

• A secondary barrier was around perceived stigma with 62% of ENR participants agreeing that they ‘do not like asking for benefits’. Only 16% agreed that their family / friends would have a negative view of them claiming PC. However, the majority of ENR participants (92%) said they would apply for PC if they knew they were entitled.

PC can offer integral financial support to pensioners in poverty, and it is essential that those eligible to the benefit receive the money they are entitled to. DWP has been investigating different ways of increasing take-up, including a ‘raising awareness’ communications campaign since April 2022. The test and learn exercise which is the subject of this report has used a new approach that identifies households who, based on our data, appear to be entitled to PC and then targets them via a letter inviting them to make a claim. This work also has wide support from parliament and stakeholders for this customer group.

1.2    Research aims

The aims of this exercise are:

1. To find out if HB data can be used to accurately identify PC ENRs.

2. To test if sending invitation to claim letters to these households prompts them to claim PC.

3. To explore claimants’ understanding of the letter and find out why they decided to claim or not to claim PC.

1.3    Methodology

1.3.1        Data used in the test and learn

The PC take-up statistics are based on survey data, which covers less than 0.1% of the pensioner population. It is therefore not a suitable data source for identifying individual households and contacting potential PC ENRs. However, DWP administrative benefit data can be used to identify a subset of all potential PC ENRs.

A dataset containing HB and PC data from the 30th June 2023 was used to identify a sample population. We used HB data as these data feeds contain information such as income and capital, which can help us to identify potential PC ENRs. This information is recorded during the HB claims process and is made available to DWP via a data sharing agreement with Local Authorities. The PC data is held by DWP and was used to remove any HB claimants that were already claiming PC.

The PC eligibility criteria^{[footnote 6]} was then applied to the capital and income information recorded in the dataset. This methodology identified approximately 144,500 potential PC ENRs across the whole of Great Britain. This figure is estimated to represent around one in six of all potential PC ENRs in Great Britain. It is currently not possible to identify all PC ENRs using the existing HB and PC administrative data, since not all PC ENRs are in receipt of HB.

1.3.2 Treatment and control groups

Treatment Group

From this, a sample of 2,409 households from 10 LA areas were identified as meeting the ENR criteria. All ENRs within the 10 areas were included.

The 10 LA areas were selected to ensure a geographically representative sample of urban, rural, regional, and national areas across Great Britain. These were:

• Charnwood

• Craven

• Eastbourne

• Harrow

• Pendle

• Powys

• Redcar and Cleveland

• Teignbridge

• Vale of White Horse

• West Lothian

All 2,409 pensioner households were sent an invitation to claim letter on the 17th July 2023. A follow-up letter reminding ENRs to claim was sent on the 12th September 2023. The letters that were sent can be found in at Annex A.

Control Group

The control group consists of every other pensioner household from the sample population that at the time of the study, did not reside in one of the 10 treatment areas. Households within the control group were not sent an invitation to claim letter and are treated as a business-as-usual case for purposes of comparison.

Using the original identification criteria, the control group represents around 142,000 households.

The treatment and control groups are similar based on the variables which have been used to select these households. For example, all households are in receipt of HB but not PC. While the 10 treatment areas were selected to be representative of a range of geographical demographics throughout Great Britain, they are not a true random sample of the population of interest. This means that whilst the control group provides a sense of the claiming behaviour of households who were not sent an invitation to claim letter, we cannot conclude that we would observe the same outcomes if DWP were to send the same letter to all households in the control group.

The effect of the refined criteria on sample sizes and representativeness

New data identified in early stages of the project allowed us to refine the criteria used to identify PC ENRs (see section 1.3.4 for further details), the effect of applying the refined criteria to both the treatment and control group alters the sample size of both groups. Diagram 1.1 shows the effect of applying the refined criteria to the original sample.

Diagram 1.1 The effect of the refined criteria on the treatment and control sample sizes

Original treatment group

Step 1: Original treatment group (n=2,409) sent an Invitation to Claim letter

Step 2: Refined treatment group (n=708) and Remaining households (n=1,701)

Original control group

Step 1: Original control group (n=142,100) not sent an Invitation to Claim letter

Step 2: Refined control group (n=45,400) and Remaining households (n=96,700)

Although the refined criteria reduced the size of the original groups by around two- thirds, the data remained representative. For example, geographical areas with larger numbers of ENRs in the original groups continued to have larger numbers of ENRs after the refined criteria had been applied. In fact, in terms of marital status, we found that the refined group was actually more representative of the total ENR population, as our original group had a higher proportion of couples when compared to ENRs within the FRS. This was largely due to not taking partner State Pension income into account when identifying ENRs in the original group.

The report will cover the findings of using both the original and refined criteria to identify PC ENRs. It should be noted that the results of the refined group are more reflective of the success of the exercise, as we can be more confident that households in the refined group are ENRs than households in the original group.

1.3.3 Quantitative analysis

DWP analysis identified PC ENRs within the selected LA areas, using information captured as part of the HB claim process.

The National Insurance numbers (NINO) of these individuals were used to track whether these households contacted DWP to make a claim for PC, using claims data recorded in the Pensions Customer Account Management (CAM) system.

Information on claims recorded on the online Apply for Pension Credit (AfPC) portal were also used to track subsequent claims. Successful PC claims are transferred to CAM for processing, therefore all remaining claims from AfPC are unsuccessful.

Several pieces of information must be provided by the claimant to make a claim for PC, including details of savings and other sources of income. The CAM system provides data showing whether a claim has been started and provides a date when the application process has been completed.

The CAM system also allows DWP staff to assess eligibility for PC. Households who meet the eligibility criteria for PC are transferred to the Central Payment System (CPS), which provides a second source of administrative data.

Therefore, the quantitative data collection process tracks the number of households at each stage of the process shown in the Diagram 1.2.

Diagram 1.2 The quantitative data collection process

1.3.4 Lessons learned about the ENR selection criteria

During the initial stages of the data collection and analysis, we identified new data that showed two-thirds (approximately 1,700) of the households who were sent an invitation to claim letter, were in retrospect not in fact a PC ENR. There were several reasons for this amongst the original sample, namely:

• State Pension income was not taken into account for new State Pension claims received via the Get Your State Pension service.

• Partner State Pension income was not taken into account.

• Partners of HB claimants who were already receiving PC (with the HB claimant being recorded as the partner of the PC claimant).

• Non-eligible mixed-age couples being included in the original identification criteria.

• Income from some DWP benefits and tax credits not being taken into account.

• Occupational pension income not being taken into account.

Conversely, after refining the ENR identification criteria, we were able to identify other households who could potentially have been entitled to PC. These are households:

• Who are in employment and may also be either a lone parent / disabled / carer. These households have a portion of their income disregarded when assessing PC eligibility.

• Who reached SPa prior to April 2016, and are entitled to only the SC element of PC.

• Who have an eligibility to a child addition, severe disability and / or carer premium within PC.

• Whose tariff income was overestimated. The initial methodology used a tariff income amount recorded on the HB data, which deducts £1 for every £250 over £6,000 in HB. The refined method used the actual claimant and partner capital recorded on the HB data, to determine tariff income, using the more generous PC measure of £1 for every £500 over £10,000.

• Who are “passported” to HB for “non-standard” reasons. i.e. mixed age couples in receipt of a working-age benefit, who are eligible for (and possibly would be financially better off claiming) PC.

Therefore, a significant outcome of this test and learn exercise has been to enable us to improve our ENR selection criteria from the data available. As a result, we are now more confident that those in the refined group are far more likely to be ENRs than was the case with the original group. This is evident in the increase of eligibility rates when comparing the original treatment group and the refined group which doubled from 36.4% to 75.3% and could, perhaps, increase further towards 90%^{[footnote 7]} once all claims have been given enough time to make their way through the claims process to completion/award.

Ideally, we would have been able to use the refined criteria to identify ENRs from the outset of this exercise. However, as noted in the ‘Research aims’ in section 1.2, one of the aims of this research was “To find out if HB data can be used to accurately identify PC ENRs”. This is the first time that HB data has been used for the operational purpose of identifying and contacting potential PC ENRs. This meant that despite our best efforts, further improvements were needed after the construction of the initial sample to improve its accuracy. As a result, one of the outcomes of this test and learn exercise was that we tested our ENR identification criteria on a small sample of the PC ENR population from the HB data and based on analysis of collection of early results data. From this, we learnt that with a few adjustments to the ENR identification criteria, the HB data could be used to accurately identify a subset of PC ENRs.

The quantitative results section of this report will consider the number of households who were sent a letter during the exercise, both out of the original treatment group of 2,409 households, and the 708 households who we would now identify as an ENR using the refined criteria.

1.3.5        Qualitative interviews

19 telephone interviews were conducted with individuals who had received the invitation to claim. The interviews were carried out using a topic guide setting out important areas to be covered in each interview, whilst also allowing the researchers and participants with flexibility to provide and explore relevant feedback. The interviews covered background information, the participants claim history, subsequent decisions to claim or not to claim after receiving the letter, before being asked about the content of the letter, messaging and initial reactions.

Individuals were invited to opt-in if they wished to participate in the research, or to opt-out if they did not wish to be contacted. As the desired number of participants was not reached from opt-ins alone, a small number of participants who had not opted-in or opted-out received a call enquiring if they would like to take part.

Participation in the interviews was completely voluntary, and participants were reminded that they could withdraw from the interview at any time.

Participants from a range of age bands (66 to 90+) were invited to take part in this research. It is worth noting that all interview participants were under the age of 85, with a skew towards younger pensioners deciding to take part.

Interviews lasted approximately 30 minutes and were conducted over the telephone at times agreed with participants. All interviews were recorded using manual notes, with the participant’s consent. Interviewers were briefed and mindful of the considerations associated with interviewing elderly participants. Assurances of confidentiality were given at the start of each interview. All participants were offered a £20 shopping voucher as a thank you for taking part.

1.3.6 Ethical considerations

We engaged with the DWP Ethics Advisory Panel to ensure that this research met the standards for free and informed consent; confidentiality and anonymity, and mitigating for the risk of distress for either participant or interviewer. All data has been handled and stored in line with GDPR and DWP governance.

1.4 Reading the report

During the report PC claimants are referred to as ‘Claimants’. Some participants in the qualitative interviews did not make a claim for PC, therefore when discussing the qualitative results, the term ‘participants’ is used when referring to this group more generally.

For some interview questions a small proportion of claimants said they did not know the answer or did not answer the question. We have not highlighted these results in the commentary unless they represented a signification number of participants.

Verbatim quotations are used throughout the report to illuminate and bring to life key findings and are attributed as follows: “Quotation.” (Age band, successful / unsuccessful claim).

Diagrams are used to illustrate quantitative results.

2.        Quantitative results

In the following analysis we will refer to two groups of households who were part of the test and learn exercise:

1. “Original treatment group”: the group of 2,409 households in the 10 LA areas who were initially identified as the group of PC ENRs living in these areas.

2. “Refined treatment group”: administrative data analysed during the exercise revealed that 708 of the original sample were in fact ENRs. The refined sample of 708 are a subset of the original sample of 2,409.

1,701 pensioner households were included in the original sample but not the refined sample. New data that became available during the exercise revealed that we can now be confident that these households would not have been entitled to PC at the time that the letter was issued. Despite this, a small number of these households had a change of circumstances since the letter was sent to them. This change of circumstance meant that when they made a claim for PC, they were eligible. There were 72 households for which this was the case, these claims are included in the overall claim success rate results.

Diagram 2.1 Pensioner households within the original treatment group who had a change of circumstance, making them eligible for Pension Credit

Step 1: Original treatment group (n=2,409)

Step 2 : Refined treatment group (n=708) and Remaining households (n=1,701)

Step 3: Remaining households - households who had a change in circumstances since the letters were sent and made a successful claims (n=72)

2.1    Response rates to the invitation letter

Refined group:

708 pensioner households were in the refined treatment group. Of these:

• 259 (36.6%) subsequently made a claim to PC during the almost 4-month period following the mailing of the letter.

The comparative figure for the refined control group was 6.2%. This is a significant result and means that those in the refined treatment group who received the ‘Invitation to Claim’ letter were 6 times more likely to make a claim than those in the refined control group who did not receive a letter.

Diagram 2.2 Claims made in the refined treatment group compared to the refined control group

Step 1: Refined treatment group

Step 2: Either ‘Sent an Invitation to Claim’ letter (n=708) or ‘Not sent an Invitation to Claim’ letter (n=45,400)

Step 3: From ‘Sent an invitation to claim’, ‘Made a claim’ (n=259) = 36.6% response rate, or

Step 3: From ‘Not sent an Invitation to claim’, ‘Made a claim’ (n=2,800) = 6.2% response rate

Original group:

2,409 pensioner households received an invitation to claim letter, this is the original treatment group. Of these:

• 713 (29.6%) subsequently made a claim to PC during the almost 4-month period following the mailing of the letter.

The comparative figure for the original control group was 3.8%.

Response rates in both treatment groups were higher than the control groups. This suggests that households that were sent a letter were more likely to make a claim than those in the control groups, who received no intervention. Between the original treatment and control groups, there were an additional 25.8% of PC claims amongst households that were sent a letter.

It should be noted that only the NINO of the HB claimant was tracked onto the PC claims data. This means that in couples, where the HB claimants’ partner made the PC claim (and the claim is subsequently found to be unsuccessful), we will be undercounting the response rate.

2.3    Eligibility rates among those who made a claim

Refined group:

Of the 259 pensioner households in the refined treatment group who made a claim:

• 195 (75.3%) were entitled to PC.

The comparative figure for the refined control group was 75.0%.

This means that 27.5% of those that DWP wrote to in the refined sample subsequently made a successful PC claim. The comparative figure for the refined control group was 4.6%.

It is clear to see, therefore, that sending an ‘Invitation to Claim’ letter had a significant effect on successful claim rates, with those in the refined treatment group being 6 times more likely to make a successful PC claim than those in the refined control group who did not receive a letter. This result is on par with other mailing initiatives that we are aware of, such as the recent initiative funded by the Mayor of London, and carried out by Policy in Practice (PiP), which saw a PC claim success rate of around 28%^{[footnote 8]}.

The effect of the letter is even more pronounced when looking at the results for those entitled to smaller amounts of PC. Compared to the refined control group, those entitled to less than £10 per week of PC in the refined treatment group were 14 times more likely to make a successful claim following receipt of the letters. This number reduces as the entitled amount of PC increases, down to just over 3 times more likely (22% v 7%) to make a successful claim for those who were entitled to £100 and over per week in PC.

This too is a significant result, as we would expect it to be more difficult to convince those who are entitled to smaller amounts of PC to make a claim, due to the lower financial reward that they would receive. For the refined treatment group, however, these results appear to show the opposite, with the letter having a bigger impact on those entitled to smaller amounts.

This trend is also apparent in the refined control group. As well as being 6 times less likely to make a successful claim to PC, the claims success likelihood of those who did not receive a letter decreased as the amount that they were entitled to decreased (down to 2% for those entitled to less than £10 per week of PC). This confirms that without the receipt of the letter, those entitled to smaller amounts of PC are less likely to make a successful PC claim. This pattern is not replicated in the refined treatment group, with no reduction in the likelihood of claims success as the entitled amount of PC reduces.

Diagram 2.3 Successful claims in the refined treatment group compared to the refined control group

Step 1: Refined group

Step 2: Refined treatment group: Made a claim (n=259) or Refined control group: Made a claim (n=2,800)

Step 3: Refined treatment group: successful claims (n=195) = 75.3% eligibility rate

or

Step 3: Refined control group: successful claims (n=2,099) = 75% eligibility rate

Original group:

DWP assessed the eligibility of the 713 pensioner households in the original treatment group who made a claim, and found that:

• 267 (37.4%) were entitled to PC.

The comparative figure for the original control group was 61.7%. This result is unsurprising given the significant number of ineligible households in the original sample.

This means that 11.1% (267) of the 2,409 DWP wrote to subsequently made a successful PC claim. The comparative figure for the original control group was 2.3%.

Although our data suggests that all of those in the refined group may be entitled to PC, we will never quite reach 100% eligibility amongst this group. This is mainly due to:

• The length of the claims process, as a proportion of more recent claims included in the analysis have not been given enough time to go through the claim process and be fully assessed for eligibility.

• Claims which are currently classed as “not entitled” due to either the application being withdrawn or DWP awaiting further documentation from the claimant etc.

The claims covered in this analysis cover a period from the mailing of the letters to approximately 4-months after this event. It is possible that in the weeks and months following this period that some of these claims would have progressed through the claims process, resulting in a successful claim. The effect of this is that the eligibility rate would be higher than reported above (perhaps heading towards 90%). However, as there are several factors that lead individuals to claim, there is a risk that extending the analysis period reduces the likelihood of a subsequent claim being linked to the invitation to claim letter. For example, future claims within the original or refined treatment groups could arise from a recent change of circumstance, a recent communications campaign etc. Additionally, there will always be a residual number of claims that are deemed ineligible (such as those who subsequently withdraw their PC application), meaning we are unlikely to reach absolute eligibility.

2.4    Average weekly amount of Pension Credit entitlement

The mean average weekly PC claimed by the 267 pensioner households in the original treatment group who made a successful claim was just over £46. This increased to just under £51 per week for the 195 pensioner households in the refined treatment group.

Diagram 2.4 Mean awards by Local Authority area in the original treatment group and aggregated groups

Local authority area	Mean awards
Pendle	£20.47
Craven	£33.57
Vale of White Horse	£36.23
Redcar and Cleveland	£44.49
Charnwood	£44.62
West Lothian	£47.54
Eastbourne	£48.87
Powys	£49.36
Teignbridge	£51.18
Harrow	£58.46
Original treatment group	£46.44
Original control group	£63.23
Refined treatment group	£50.57
Refined control group	£67.24

The mean average award was highest in Harrow (£58), whilst the lowest mean award was in Pendle (£20). The mean average awards are higher in the original control and refined control areas, this may be because claimants in the control groups are more inclined to claim if they are confident that they are entitled to a larger amount of PC per week.

2.5    Age group of those eligible

The mean age for a successful claimant within the original treatment group was 72 years old, this fell to 71 years old for the refined treatment group. The same ages are found in the original control group (72 years old), and the refined control group (71 years old). This shows that the successful claims are not skewed by an influx of claimants recently reaching SPa.

Diagram 2.5 Mean ages of successful claim by Local Authority area in the original treatment group and aggregated groups

Local authority area	Mean ages
Powys	69
Pendle	70
Vale of White Horse	71
Eastbourne	71
West Lothian	71
Teignbridge	72
Harrow	72
Craven	73
Charnwood	74
Redcar and Cleveland	74
Original treatment group	72
Original control group	72
Refined treatment group	71
Refined control group	71

Powys has the lowest mean age in the original treatment group, with successful claimants averaging 69 years old. Meanwhile both Charnwood and Redcar and Cleveland had the highest mean age in the original treatment group, with successful claimants averaging 74 years old.

Diagram 2.6 Median ages of successful claim by Local Authority area in the original treatment group and aggregated groups

Local authority area	Median ages
Powys	69
Eastbourne	70
Harrow	70
Teignbridge	70
Vale of White Horse	70
Pendle	71
West Lothian	71
Charnwood	72
Redcar and Cleveland	72
Craven	74
Original treatment group	71
Original control group	70
Refined treatment group	70
Refined control group	69

Similar positives can be taken from the median age analysis, with successful claimants within the original treatment group and refined treatment group being at least 4 years above the SPa.

There are no significant differences between the mean and median ages of the successful claimants and unsuccessful claimants.

2.6    Gender of those eligible

We see large amounts of variation in the gender breakdowns within our samples, largely owed to sample sizes.

Diagram 2.7 Gender breakdowns for successful claimants by Local Authority area in the original treatment group and aggregated groups

Local authority area	Female	Male
Charnwood	58%	42%
Craven	55%	45%
Eastbourne	45%	55%
Harrow	55%	45%
Pendle	57%	43%
Powys	61%	39%
Redcar and Cleveland	43%	57%
Teignbridge	38%	62%
Vale of White Horse	42%	58%
West Lothian	62%	38%
Original treatment group	52%	48%
Original control group	53%	47%
Refined treatment group	49%	51%
Refined control group	53%	47%

Within the original treatment group, 52% were female and 48% male. For the original control group sample these change to 53% and 47%, respectively. These results are in line with the incumbent caseload.

2.7    Comparison of treatment and control areas

The below chart depicts the differences in response rates (RRs) and entitlement rates (ERs) across LA areas within the treatment groups and how these rates compare to the control groups. The orange bars depict the rates using the original criteria and the grey bars show the improvement in rates if we use the refined criteria. The data labels show the sum of the two. For example, in Charnwood 31% responded in the original treatment group, and 41% responded within the refined treatment group. The grey bar depicts the additional 10% gain in the response rate when analysing the refined group.

Diagram 2.8 Response rates and entitlement rates by Local Authority in the original treatment group

Local authority area	Original criteria estimates	Refined group estimates added
Charnwood (response rate)	31%	41%
Charnwood (entitlement rate)	47%	96%
Craven (response rate)	28%	34%
Craven (entitlement rate)	37%	80%
Eastbourne (response rate)	27%	28%
Eastbourne (entitlement rate)	21%	50%
Harrow (response rate)	32%	33%
Harrow (entitlement rate)	44%	71%
Pendle (response rate)	22%	25%
Pendle (entitlement rate)	30%	71%
Powys (response rate)	25%	33%
Powys (entitlement rate)	34%	84%
Redcar and Cleveland (response rate)	33%	46%
Redcar and Cleveland (entitlement rate)	32%	69%
Teignbridge (response rate)	33%	42%
Teignbridge (entitlement rate)	31%	70%
Vale of White Horse (response rate)	30%	38%
Vale of White Horse (entitlement rate)	39%	80%
West Lothian (response rate)	27%	36%
West Lothian (entitlement rate)	48%	80%
Treatment group (response rate)	30%	37%
Treatment group (entitlement rate)	37%	75%
Control group (response rate)	4%	6%
Control group (entitlement rate)	62%	75%

2.8    Reduction in ENRs and amount of unclaimed PC

As mentioned above, of those who were sent an invitation to claim letter, 267 resulted in a successful claim to PC, with a weekly mean average award of just over

£46. This equates to around £0.65m being removed from the amount of PC that is unclaimed each year^{[footnote 9]}.

A proportion of these households may have claimed PC under normal circumstances (i.e., without a letter prompt). Assuming the same response and eligibility rates from the original control group, and applying them to the original treatment group, we estimate the letter has resulted in an additional:

• 211 (79%) to successfully claim PC

• £0.51 m (79%) being removed from the annual unclaimed PC; than would have been the case had they not received the letter^{[footnote 10]}.

2.9    Limitations and assumptions

1. Some of the original sample of households who received an invitation to claim may have died since the letters were mailed. We have not subtracted deceased claimants from either the numerator or denominator of our response rate or eligibility rate calculations.

2. It has been assumed that claims made after the invitation letter would have been submitted within 4 months of the letter being sent. Any individual who made a claim for and received PC after 17th November 2023 is not included in our response rate or eligibility rate calculations.

3.      Qualitative results

The following section reports on findings from 19 telephone interviews conducted with recipients of the invitation to claim letter.

15 of the 19 of participants said they made a subsequent claim after receiving the letter.

Of the 4 remaining participants, 2 said they had made a claim for PC in the past 12 months but before receiving the letter. As both participants stated this claim was made prior to the invitation to claim, they cannot be attributed to the intervention and have not been treated as such. This means that only 2 participants did not make a claim following the invitation to claim letter for reasons other than having a claim assessed within the past 12 months. Due to this small response rate, this report does not cover reasons why participants did not claim or barriers to claiming.

3.1    Awareness of Pension Credit and claim history

All participants had heard of PC before. The most common places participants had heard of PC was from friends, family, or they had experience of claiming either for themselves or with their partner in the past. Most participants understood PC to be a top-up benefit for pensioners to boost their income. Some mentioned that PC eligibility can get you extra help or benefits such as help with glasses or dental care.

The majority said they had experience of making a claim for PC sometime in the past. Some said that they had never made a claim before. All participants who had never claimed before made a claim following the letter, and all of these claims resulted in successful awards.

There was a common perception amongst participants that they would not be entitled to PC, even amongst those who claimed successfully. Many participants who had claimed in the past and were assessed as ineligible assumed they would still not be entitled, whilst others thought their income was too high to be entitled to PC. A small number of participants were shocked that they could claim again, with some also not being aware that they could claim PC alongside other benefits.

3.2    Understanding of the Invitation to Claim

Generally, understanding of the invitation to claim letter was good. Most participants understood that the letter meant they were potentially entitled to PC. Some interpreted the letter as meaning they were definitely entitled, potentially contributing to a heightened sense of frustration if the claimant was subsequently assessed as ineligible.

Most participants gave positive feedback with regards to the content and structure of the letter. Participants said that the letter was clear, straightforward, and informative.

Most participants understood the purpose of the letter as boosting claims or encouraging them or individuals like them to claim money that they were potentially entitled to. Most thought that it was good that DWP were conducting the activity to encourage ENRs to claim. Even those who were not entitled could see the positive aims of the activity, despite not receiving the outcome they hoped for.

Worthwhile exercise that the DWP were trying to make sure we got all the benefits we were entitled to” (80-84, Unsuccessful claim).

It encouraged me to have a go basically […] and it may have helped other people find out they were eligible for it” (65-69, Unsuccessful claim).

3.3    Motivations to Claim

Most participants said they made a claim after receiving the letter. Of those 15 who did make a claim, 8 were successful and 7 were unsuccessful.

The majority of participants cited the invitation to claim as their primary motivation to claim. Of those that were asked, all participants said that the letters had influenced their decision to make a claim. Although this research cannot establish causation, the qualitative feedback supports the higher response rates among households in the treatment areas compared to the control groups. Some claimants said of their own accord that they would not have made a claim if they had not received the letters.

I wouldn’t have made a claim if I didn’t get them [the letters]” (65-69, Unsuccessful claim).

Claimants were also asked of any other reasons they made a claim, allowing two main secondary motivations to be identified:

• Participants said that it was worth a try. Claimants who gave this reason made a claim just in case or to see if they were entitled. These participants viewed making a claim as a chance for some extra financial help. As one claimant said:

I’m not going to lose anything, but I might gain” (65-69, Successful claim).

• Financial difficulty and struggling to cope with the cost of living also arose as a secondary reason to claim. Some participants said they were just able to get by on their current level of income, while others mentioned the financial pressure of bills. For these participants, an award of PC would be welcomed to support their living costs.

3.4    Reaction to the Invitation to Claim

Participants’ reactions to the invitation to claim process fell into two main themes, either being categorised as positive or mildly frustrated. In some cases, reactions appeared to depend on the participant’s claim outcome, with claimants who received successful awards more likely to respond positively, and claimants who made unsuccessful claims more likely to express frustration.

Those that were categorised as positive expressed pleasant surprise in being contacted as they often did not think that they would be eligible for any more support than they were currently receiving. Those who were frustrated felt disappointed that they were invited to make a claim and subsequently assessed as ineligible.

Some claimants said the follow up letter served as a good reminder, either reminding them to act or prompting them to consider claiming more seriously. However, for a subset of those who had claimed unsuccessfully, the follow up letter served as a further topic of frustration.

Some participants also expressed a reaction of surprise or shock that they received the letter, potentially linked to a wide perception amongst the cohort that they were not entitled to PC.

3.5   The Effect of Pension Credit

The topic of the cost of living arose as an emergent theme in most interviews, both with claimants and non-claimants.

Amongst those who claimed successfully, some mentioned the positive effect of being awarded PC. One participant who spoke about their bills piling up with the cost of living said that the successful PC award “pulled us back from the edge again” (70- 74, Successful claim). Meanwhile, another claimant said the influence of being entitled to PC meant that they could put the heating on in the evening during the winter months (65-69, Successful claim).

4.      Conclusions and implications

The results from the exercise show that a significantly higher proportion of the treatment group made a successful claim to PC, compared to the control group. This is particularly the case within the refined groups (27.5% treatment v 4.6% control).

This strongly suggests that the invitation to claim letter had a positive effect on prompting pensioner households to make a claim for PC, with those receiving the letter being 6 times more likely to make a successful claim than those who did not receive the letter. The effect of the letter is even more pronounced when we look at the results for those entitled to smaller amounts of PC, with comparative findings between the refined control and refined treatment groups showing the smaller the entitled amount, the bigger the impact the letter had in increasing the likelihood of a successful claim to PC.

Findings from the qualitative interviews support this conclusion, as most participants felt that the letters prompted them to claim. The invitation to claim letters emerged as the primary motivation to claim, with secondary motivations including it being ‘worth a try’ and struggling with the cost of living. Broadly, understanding of the messaging and purpose of the letters was good, and participants appreciated DWP contacting them to see if they were eligible.

A significant outcome from the test and learn exercise is that it has allowed us to refine the criteria we could use to identify PC ENRs using HB data. For example, in the refined treatment sample, this increased to 75.3% of those who claimed being successfully awarded PC. The average mean award for this group is £51 per week, to support pensioners with the cost of living plus the aid of other passported benefits.

The refined ENR identification criteria identified just under 121,000 HB claimants who may be entitled to PC but not receiving it. If we assume that the results from the refined group can be replicated throughout Great Britain, we estimate that this could lead to an additional 28-35% (32,000 to 41,000 claimants) making a successful claim to PC. This would lead to an overall increase in PC take-up, a reduction in the annual unclaimed PC, and more pensioner households receiving the support they are entitled to.

1. Income-related benefits: estimates of take-up: financial year ending 2022 - GOV.UK (www.gov.uk) ↩

2. Around one in six PC ENRs claim HB ↩

3. There were 267 households from the original treatment group that made a successful claim to PC; this is higher than the 195 successful claims in the refined treatment group. These 72 households were not included in the refined treatment group. This is because based on the refined criteria, they were not entitled to PC at the start of the exercise. In the period following the letters these households had a change of circumstance, meaning that when they made a claim for PC they were subsequently entitled. These 72 successful claims have been counted towards the original treatment group. This is explained further in the quantitative results section of the report. ↩

4. Income-related benefits: estimates of take-up: financial year ending 2022 - GOV.UK (www.gov.uk) ↩

5. Pension Credit eligible non-recipients: Barriers to claiming (publishing.service.gov.uk) ↩

6. In simple terms, if the income for a single / couple household is below the Standard Minimum Guarantee rate (£182.60 for singles, £278.70 for couples in 2022/23), then they are entitled to Pension Credit. ↩

7. The reasons for this increase towards 90% are outlined in more detail in section 2.3. ↩

8. How London used data to make pensioners £8 million a year better off ↩

9. The latest PC take-up statistics estimate that for the financial year ending 2022, the amount of unclaimed PC was up to £2.1bn. Income-related benefits: estimates of take-up: financial year ending 2022 - GOV.UK (www.gov.uk). ↩

10. These findings relate to the additional number of claims and amount of PC claimed as part of the original treatment group. We have used the original treatment group to reflect the claims relating to the 72 households who received a letter as part of the trial and, due to a subsequent change in circumstances, were found to be eligible for PC. ↩